Edited images: the proof

Questions, speculations & updates on the techniques and nature of media fakery
burningame
Member
Posts: 109
Joined: Thu Dec 09, 2010 1:21 pm

Re: Edited images: Any proof at all?

Unread post by burningame »

Well, that is exactly what I want to know myself! I was extrapolating from what Krawetz implies.

But, I do not know - that is why I'm asking the experts on this forum the questions in my posts. I am trying to determine if what Krawetz says about photos is true.

If you read Krawetz's blog though, to me it looks like the information used ('bitstream') is a mathematical representation of the image - I don't see how you could, after the fact, change the maths without changing the image.
fred
Banned
Posts: 592
Joined: Tue Oct 20, 2009 12:43 pm
Contact:

Re: Edited images: Any proof at all?

Unread post by fred »

Burningame and Anon, these are interesting questions.

None of the mathematical tools or tests are magical all by themselves.

Suppose that I, as a manipulator, take an image file and decide to deliberately flip one of the least significant bits--suppose I decided to make the tip of one blade of grass in a photo of a golf course a slightly lighter shade of green. Would anyone ever know? Probably not.

On the other hand, suppose I take a picture of Nancy Reagan and Tiger Woods, and merge that into the original picture. Would there be signs of tampering? Almost certainly.

---

How about some analogies with other forms of tampering?

Suppose a disgruntled restaurant worker decides to tamper with food. If he were to put one single molecule of bacteria into your soup, you'd probably never find out about it and neither would a very sophisticated laboratory because there are millions of bacteria floating around in the air all the time, and so finding one molecule in the soup doesn't suggest tampering. If this same employee decided to empty a whole bottle of rat poison into the soup, the odds are very high that the tampering would be discovered. Maybe the "normal level" of arsenic in soup is one part per billion, and in the doctored soup it's a million times higher than that. At that point the soup would probably smell funny and taste funny. Even without a lab you might notice that there's something very wrong with the soup.

Mose of the 9/11 images are like the soup with the whole bottle of rat poison. The toxicology tests merely confirm what an astute observer can see with his own eyes. Shadows and illumination are all wrong, buildings are out of place, the laws of physics are consistently violated, sections of smoke move in reverse, devil faces appear in the smoke, the soundtrack is hideously bad.... the list goes on and on.

While you're waiting for those toxicology reports to come back from the lab, heed some friendly advice: don't eat the soup.

-----

The computers are handy for detecting certain types of forgery that is less apparent to the naked eye, such as copy-move forgery. By way of further analogy, suppose that I hand you a printed list of every name and telephone number in Chicago. It's not that easy to figure out by hand if there are 20 different people with the same telephone number in that list, but it's a trivial task for the computer.

To the naked eye, one section of blank wall or blue sky looks pretty much like another. To the computer, it's easy to spot that "this section of blank wall EXACTLY MATCHES that other section of blank wall, therefore something's wrong with this picture!"

Ultimately people have to decide for themselves whether or not something is fake or real, the computer just helps to automate otherwise dreary tasks and to point out areas of likely image manipulation.

-----

If I crack open a rotten egg I don't have to actually eat the whole thing, or ask a PhD in Food Science for his opinion and send a few samples to a independent laboratories in order to know that the egg is rotten. When somebody comes into the store with a bag full of fresh $50 bills hot off the printer all with the same serial number, I don't need to wait for the Secret Service to arrive and tell me it's counterfeit.

----

All of the hand-wringing over never being able to know anything without fill-in-the-blank is really a poor excuse for the fake crap published by the lying media perps and their army of semi-retarded shills.

Disclaimer: I've never been to the moon so I cannot comment on moon frogs hopping past NASA astronauts, but I can tell you those WTC 9/11 videos are fake.
burningame
Member
Posts: 109
Joined: Thu Dec 09, 2010 1:21 pm

Re: Edited images: Any proof at all?

Unread post by burningame »

Fred wrote:

...[most] of the 9/11 images are like the soup with the whole bottle of rat poison. The toxicology tests merely confirm what an astute observer can see with his own eyes. Shadows and illumination are all wrong, buildings are out of place, the laws of physics are consistently violated, sections of smoke move in reverse, devil faces appear in the smoke, the soundtrack is hideously bad.... the list goes on and on...

...To the naked eye, one section of blank wall or blue sky looks pretty much like another. To the computer, it's easy to spot that "this section of blank wall EXACTLY MATCHES that other section of blank wall, therefore something's wrong with this picture!..."
Well then, that’s great! But why, then, do we not have these scientific proofs in our arsenal? If it’s akin to a whole bottle of rat poison, it should be fairly easy to confirm with forensic tests, right?
anon1911
Banned
Posts: 61
Joined: Mon Jul 25, 2011 2:57 pm

Re: Edited images: Any proof at all?

Unread post by anon1911 »

Thank you for these very contributive words, fred. I fully agree with most of your analogies.
fred wrote: To the naked eye, one section of blank wall or blue sky looks pretty much like another. To the computer, it's easy to spot that "this section of blank wall EXACTLY MATCHES that other section of blank wall, therefore something's wrong with this picture!"
That's interesting. Let's check out some example from the Oslo thread. Someone (sorry, forgot his name) just recently linked to a software which is able to detect 'copymove' parts of the picture. I think this is actually a tool that is really helpful if used correctly.

Image
Here's the original.

Image
Here's the image in which the above mentioned software detects copymove parts. Well, what is your opinion about this?
My opinion is, that the software is correct, but not because the image has been tampered with but because the wall behind that guy is pretty much blured and washed out. Some small parts repeat each other because the wall does not have a certain texture.

I'll give this some forensic taste. I claim the following:
The bigger the parts of the image that are analysed and compared with each other, the higher the probability that it really has been copied on purpose if they fit. The other way around of course is: The smaller the parts, the higher the risk that you find random matches.

If you analyse 16x16 (pixel) parts of an image and you want to get noticed whenever more than 70% of these pixels fit to each other, you'll maybe get reliable results.
If you analyze 4x4 (pixel) parts, the possibility is way higher to have random matches.
If I crack open a rotten egg I don't have to actually eat the whole thing, or ask a PhD in Food Science for his opinion and send a few samples to a independent laboratories in order to know that the egg is rotten.
Thanks for that analogy! It's so true but still made my day :lol:

If it’s akin to a whole bottle of rat poison, it should be fairly easy to confirm with forensic tests, right?
First we have to determine which forensic analysis serve our needs and if they are reliable at all.
Last edited by anon1911 on Sun Sep 11, 2011 4:15 pm, edited 1 time in total.
fred
Banned
Posts: 592
Joined: Tue Oct 20, 2009 12:43 pm
Contact:

Re: Edited images: Any proof at all?

Unread post by fred »

Burningame wrote:Well then, that’s great! But why, then, do we not have these scientific proofs in our arsenal? If it’s akin to a whole bottle of rat poison, it should be fairly easy to confirm with forensic tests, right?
You're right. We absolutely should have the latest and greatest forensic tools in our arsenal. Fortunately, BrianV is working on it.

One of the problems we've run into in the past is that the money for developing these tools tends to be all on the "other side". The NSA and NRO and NGIA and NASA have many billions of dollars to hire graduate students and programmers and develop tool kits for the Empire. As the "Galactic Rebels" we're limited to what we can get our hands on. The professors who work on image analysis tend to be already on the payroll of the Death Star and completely dependent on the Military Industrial Complex for grant funding, promotions, etc. So they're not overly enthusiastic to expose their own side's mistakes.

We should really do a recruiting trip up to MIT and get some of the kids in their Media Lab working on this stuff.

There are a lot of biometric tools that would be very useful to us and that have been on my wish list for a long time, especially facial recognition. We could then show that "Vicsim A" has the same chin as "Vicsim B" and the same eyes as Vicsims "C", "D", "E", and "Q". I'm sure with better tools we would find many additional glaring examples of fraud.

As time goes by more and more talented people start working on these problems of exposing psyops and hoaxes they bring their skills and tools with them. It took a long time from NASA's moon hoax until regular people could store and analyze video on computers (thirty-plus years?). If you wanted a video tape of the moon landings you had to wait for VCR's to become available to consumers and then find someone with the tape who would sell it to you. Now we have dozens of independent people exposing NASA's Mickey Mouse efforts on youtube and it takes only minutes to obtain the videos. The lead time is getting shorter and shorter. For 9/11 it took five or ten years to put together convincing evidence of video fakery, and now with the Oslo psyop things are getting exposed within weeks.

Strangely, the video fraud doesn't seem to be keeping pace with the technology. I think that the "bad guys" have decided that trying to make undetectable fraud is a fool's errand and that they're just going for the mass of public opinion of people who don't think very much. There are better tools available to the perps than copy-move forgery, but copy-move forgery is easy so they rely on that heavily. They're pretty lazy and they go with what has worked for them on previous operations.

It's like NASA deciding not to include stars in their pretty pictures. If anybody has astronomers on the payroll it's NASA. But getting the stars right in every picture would be a lot of work and would slow them down, so it's easier just to ignore the issue altogether. If someone complains about a lack of stars, they just say "Hey, we're NASA. You're not an astronaut. You don't understand space pictures. Screw you!"

We see from the halfhearted infiltration attempts here that the perps know they're never going to win the battle against intelligent people. They're putting their hopes behind the idea that most people are too scared to investigate these matters for fear of being labeled a "conspiracy theorist". Rather than hire a couple of astronomers to help them get the stars in the right place and put out fewer but better images, they're willing to hire hundreds of useful idiot shills to ridicule anyone who points out their crappy mistakes and to troll message boards.

Given that the perps' primary goal is to manipulate the weak-minded general public and influence public opinion, when we do roll out new tools and offer more convincing proofs, their line of attack will be exactly the same. They'll say that the new tools don't prove anything conclusively, that there's no evidence, that we're a bunch of conspiracy tards, and that it doesn't matter if the computer says that the video is a poorly-created cartoon montage patched together from a dozen different sources, because their sister's aunt crashed into the WTC on 9/11. They'll pay Purdue University $500 to lend its name to some crappy article that claims that the video analysis tools were applied improperly. They'll trot out an 11 year old girl who was born on 9/11 to say that she still remembers the second plane crashing and hire some actress to cry at Gabby Gifford's brain transplant memorial hospital.

Ultimately, the perps are fighting a losing battle, and they know it. The Euro is falling apart and they're running psyops to try to get Norway into the EU. They're not going to get their grandiose New World Order, and they're wasting all of their capital on bullshit.
brianv
Member
Posts: 3971
Joined: Sun Oct 18, 2009 10:19 pm
Contact:

Re: Edited images: Any proof at all?

Unread post by brianv »

I don't want to get entirely hung-up on this but...

"My opinion is, that the software is correct, but not because the image has been tampered with but because the wall behind that guy is pretty much blured and washed out."

Did you consider that "blurred/washed-out" is the effect desired by the shooopers and created by copymove/cloning? Why would you put a blurred and washed-out photo of yourself on a website promoting well...yourself?
anon1911
Banned
Posts: 61
Joined: Mon Jul 25, 2011 2:57 pm

Re: Edited images: Any proof at all?

Unread post by anon1911 »

brianv wrote: "My opinion is, that the software is correct, but not because the image has been tampered with but because the wall behind that guy is pretty much blured and washed out."

Did you consider that that is the effect created by copymove/cloning?
When looking on the example I posted above you can literally see huge pixel fragments without any zoom. These small pixel fragments don't have any detail and due to the lack of detail they probably are detected by the software. I don't think that the copymove/cloning procedure does blur or wash out anything, expect on the edges around the cloned material to make it look 'real' of course.

But the blur on this specific example can be seen on the picture overall. Probably caused by the low resolution.

Edit: I just saw you edited your post too. So I'll continue:
I don't know the source of this picture where it was originally published and I don't know wheter this image was in 'high' resolution when it was published. Who would publish such a potograph of oneself? To be honest I can't answer that question.

So, if this fake industry fakes images it of course will often publish low resolution material since it is easier to fake because you don't have to care about the smallest details. Maybe it is wanted by the shoppers, maybe not. We have to find other ways to proove the media fakery I guess.
brianv
Member
Posts: 3971
Joined: Sun Oct 18, 2009 10:19 pm
Contact:

Re: Edited images: Any proof at all?

Unread post by brianv »

anon1911 wrote:
brianv wrote: "My opinion is, that the software is correct, but not because the image has been tampered with but because the wall behind that guy is pretty much blured and washed out."

Did you consider that that is the effect created by copymove/cloning?
When looking on the example I posted above you can literally see huge pixel fragments without any zoom. These small pixel fragments don't have any detail and due to the lack of detail they probably are detected by the software. I don't think that the copymove/cloning procedure does blur or wash out anything, expect on the edges around the cloned material to make it look 'real' of course.

But the blur on this specific example can be seen on the picture overall. Probably caused by the low resolution.

Edit: I just saw you edited your post too. So I'll continue:
I don't know the source of this picture where it was originally published and I don't know wheter this image was in 'high' resolution when it was published. To be honest I can't answer that question.

So, if this fake industry fakes images it of course will often publish low resolution material since it is easier to fake because you don't have to care about the smallest details. Maybe it is wanted by the shoppers, maybe not. We have to find other ways to proove the media fakery I guess.
Care to offer another example of a similarly blurred and washed-out photograph of low resolution as you describe and I will run it through!
anon1911
Banned
Posts: 61
Joined: Mon Jul 25, 2011 2:57 pm

Re: Edited images: Any proof at all?

Unread post by anon1911 »

brianv wrote:Care to offer another example of a similarly blurred and washed-out photograph of low resolution as you describe and I will run it through!
Okay, I just don't have a photograph taken by me. If anyone has a non-edited low resolution picture please provide it else I will try to get one asap.

Until then you can try the software using this picture. I'd be interested in the result.
Image
Source: http://virtual.yosemite.cc.ca.us/ghayes ... _Maree.jpg

It's not taken by me, just a random image found at google images. The sky seems to be similar washed out like the on the portrait picture from my previous example.
Note: Of course I can't proove wheter this image is fake or not but let's just assume it's not until we have one or more images provided by forum users.
brianv
Member
Posts: 3971
Joined: Sun Oct 18, 2009 10:19 pm
Contact:

Re: Edited images: Any proof at all?

Unread post by brianv »

working...

Image

You may very well have a point but as you say you can't verify the source! When you say fake, do you mean it hasn't been altered in some way?

I ran a similar random image last night and it came up clear!

Image

I will continue with the experiments!
Heiwa
Banned
Posts: 1062
Joined: Tue Oct 20, 2009 6:20 pm
Contact:

Re: Edited images: Any proof at all?

Unread post by Heiwa »

Pls analyze below AP photo:
Image
FILE - In this Tuesday, Sept. 11, 2001 file photo, United Airlines Flight 175 approaches the south tower of the World Trade Center in New York moments before collision, seen from the Brooklyn borough of New York. (AP Photo/ William Kratzke)
brianv
Member
Posts: 3971
Joined: Sun Oct 18, 2009 10:19 pm
Contact:

Re: Edited images: Any proof at all?

Unread post by brianv »

Heiwa wrote:Pls analyze below AP photo:
Image
FILE - In this Tuesday, Sept. 11, 2001 file photo, United Airlines Flight 175 approaches the south tower of the World Trade Center in New York moments before collision, seen from the Brooklyn borough of New York. (AP Photo/ William Kratzke)
Image

Thanks for the challenges!

http://lemonodor.com/archives/2008/02/p ... cally.html

I must point out this in fairness

"Unfortunately copymove isn't quite ready to be run on every AP and Reuters photo. The program takes two parameters, a “quality” (blurring) factor and a threshold. I used a quality of 10 and threshold of 20 for most of the images above, but those values don't work for all images. Some images (especially anything with a blurry background) are filled with false positives no matter what settings are used. And finally, I couldn't find an RSS or atom feed containing high quality news photos. (are there any?) "
burningame
Member
Posts: 109
Joined: Thu Dec 09, 2010 1:21 pm

Re: Edited images: Any proof at all?

Unread post by burningame »

So, judging from the previous posts I would conclude that nobody on this forum knows anything about these forensic tests, mentioned above:

Error Level Analysis
Principal Component Analysis
Color Distance Algorithm
Blur Detection
Luminance Gradient
Perceptual Hash Algorithm
Gradient Map Test

I doubt they are unique to either MIT or the perps. You'll forgive me if I don't include the 'copymove' test. Have you guys seen the Hacker Factor blog? Krawetz does more with these tests than examine the wall or the sky.

Scientific method is not a 'flavour' or a ‘taste’ that you add on; that attitude is more reminiscent of Hollywood and the MSM.

[EDITED: removed 'Color Distance Algorithm' - listed twice]
Last edited by burningame on Thu Sep 15, 2011 3:20 pm, edited 2 times in total.
reel.deal
DELETED THEIR OWN POSTS :(
Posts: 1292
Joined: Sun Aug 15, 2010 12:42 am
Contact:

Re: Edited images: Any proof at all?

Unread post by reel.deal »

.
Last edited by reel.deal on Mon Oct 01, 2012 10:17 am, edited 1 time in total.
fred
Banned
Posts: 592
Joined: Tue Oct 20, 2009 12:43 pm
Contact:

Re: Edited images: Any proof at all?

Unread post by fred »

Burningame wrote:So, judging from the previous posts I would conclude that nobody on this forum knows anything about these forensic tests, mentioned above:
How do you come up with that conclusion? Please show your work so we can all see how it follows from the posts above.
Scientific method is not a 'flavour' or a ‘taste’ that you add on; that attitude is more reminiscent of Hollywood and the MSM.
So, uh, you're going to be Mr Scientific Method around here? Maybe you should go smoke a bowl with Prof Fetzer...

Here. I saved you a couple of cut-and-pastes for later in the thread.

ImageImageImageImageImageImageImage

Judging from the posts above nobody here on the internet knows how to use an oscilloscope and a theodolite and a thermos bottle: Science this, scientific method that, epistemological anxiety, can't know this, can't know that, need an expert, very concerned, very concerned...

Time to call in Popular Mechanics and Purdue and get some very credentialed experts here to validate everything for Mythbusters and National Geographic so that we can all breathe a sigh of relief with Willy Ustad and go back to quoting the New York Times.
Post Reply